Data Rich and Information Poor
What should IoT 2.0 look like?
The phrase data rich and information poor (DRIP) was first used in the 1983 best-selling business book, In Search of Excellence, to describe organizations rich in data, but lacking the processes to produce meaningful information and create a competitive advantage. We now live in a world where IoT is exploding exponentially and large amounts of data are being collected every single day. Unfortunately, there isn’t a clear way to sort through or analyze this data within a reasonable timeframe and this phenomenon made many organizations data rich and information poor.
IoT has the power to shed light on the darkest corners of your business and, from that darkness, data can uncover countless opportunities to improve workflow processes, optimization, and ultimately drive ROI. However, like most new trends that gain initial hype for what they can ‘possibly’ do, IoT has yet to realize its true potential. Many companies and OEMs have integrated IoT into their processes and products, but are yet to realize the power that lies deep in the terabytes of data that are being collected daily.
Many of the IoT applications being deployed today are more or less focused on taking advantage of a trend rather than delivering real world solutions. Very often, these systems are difficult to justify financially. In fact, a survey by Cisco found that only 26% of IoT implementers were able to claim success for their projects. The remaining 74% of unsuccessful implementations points to dollars lost as well as a big opportunity to change how we implement IoT and use data to drive better ROI.
Seeing Apple’s success in revolutionizing consumer product UI and UX, it’s a wonder IoT developers haven’t embraced an elegant design mentality to simplify how we consume and interact with IoT data. I often ask myself and our team this question… “What if we work as an industry to move into the next phase of IoT and ensure that the data we’re collecting reaps the benefits of implementing this technology? What if we challenged the norm and forged ahead into IoT 2.0?”
If we do this and are successful, we could finally serve the right data at the right time while adopting IoT systems that reap the benefits of using the technology.
Right now, however, we are data rich and information poor. The tools and technology are there to gather massive amounts of data. Although all of this data exists in abundance, we lack the tools to analyze, optimize, and realize the true ROI that was promised when the term IoT was introduced to the world in 1999.
What needs to be done to right the IoT course and gain the insights for business process improvement? Data is useless unless it provides context and getting to this contextual state will be made easier if we begin focusing on the following…
Integrate Workflow Across the IoT Ecosystem
If your IoT efforts are not integrated with internal workflow processes overall, value and ROI is nearly impossible. Currently, most IoT systems require users to integrate workflow around the technology rather than the technology supplementing existing workflow. This is all about meeting potential customers where they are rather than requiring them to adopt an entirely new system.
For example, in machine maintenance, if you have the data on how to fix something or address an issue but the technician doesn’t know how to integrate with workflow, the data is not providing value. Your maintenance specialist is then forced to work around the data to get the job done.
Your subject matter experts (SMEs) also carry a trove of information about pieces of equipment and have their own internal workflow to ensure that things don’t go down and business runs as smoothly as possible. One day, however, these SMEs may leave and when they go, all of that information around maintenance and workflow leaves as well. To avoid this, business decision makers that are evaluating technology should attribute value to systems that integrate workflows with minimal augmentation to existing processes.
Denounce Vendor Specificity
Let’s say you have a proprietary piece of equipment and you want it to talk to another machine by a different vendor. The conversations you want your machinery to have with one another often can’t happen because the data is siloed.
Why? OEMs are in the business of making money and pleasing their investors. They want to sell you proprietary systems, maintenance plans, upgrades, and everything else in their overall product mix. Keeping their proprietary status and vendor specificity close to the chest makes machine integration nearly impossible. To that end, we need to focus on building platforms that integrate multiple IoT sources into a single application while working with the customer to make workflows part of the overall system.
At the end of the day, this is all about eliminating silos of data that are collected from hardware and building a solution that has the power to integrate multiple IoT data streams from a number of sources. These silos also exist with SMEs. Therefore, we need to make sure they have technology that is malleable enough to integrate with existing workflows. These solutions should also guide people through fixes and more while using machine learning to get smarter along the way. And of course, eliminating clunky UIs and replacing these with elegant visuals that have been gamified for ease of use is of the utmost importance. An easy (and fun) to use interface offers both spatial and intuitive context that plows through barriers to entry for things like training and data visibility in consumable formats.
Pair Automation with Machine Learning
With a goal of gaining more insight into the current state of IoT deployments, my team and I have spoken to a number of IT executives. Over and over again, they told us that they are downloading reports and conducting manual analysis to gain insights to improve processes. Why aren’t these systems automatically teeing up insights in real time? IoT enabled sensor devices have the ability to deliver real time information about the status of just about anything that happens in a day’s work. This includes the temperature of an engine on a manufacturing floor and the precise location of a piece of equipment or the time an employee is spending on a specific project.
If we can automate the delivery of real time information, why aren’t we integrating automation with machine learning so our data is more useful over time? It’s time to harness the power of machine learning in conjunction with an ongoing stream of real time data so we can easily tap into things like predictive maintenance and workforce optimization.
If we can accomplish the above, we will move away from companies being data rich and information poor to being data rich with an abundance of actionable information, ushering in a new era in IoT.
About the Author
This article was written by Angie Sticher, Co-Founder, Chief Product Officer, and Chief Operating Officer of UrsaLeo, an enterprise software company that enables users to visualize and interact with realtime operational data in a photorealistic 3D representation of their facility or equipment.